This book provides an adaptive control theory perspective on designing congestion controls for packet-switching networks. Relevant to a wide range of disciplines and industries, including the music industry, computers, image trading, and virtual groups, the text extensively discusses source oriented, or end-to-end, congestion control algorithms. The book empowers readers with clear understanding of the characteristics of packet-switching networks and their effects on system stability and performance. It provides schemes capable of controlling congestion and fairness and presents real-world applications to demonstrate the modeling and control techniques.
Author(s): Houmkozlis, Christos N.
Series: Automation and Control Engineering
Publisher: CRC Press
Year: 2012
Language: English
Pages: 332
Tags: Библиотека;Компьютерная литература;Компьютерные сети;
Front Cover......Page 1
Dedication......Page 6
Contents......Page 8
List of Figures......Page 14
List of Tables......Page 20
Preface......Page 22
1.1 Overview......Page 26
1.2 Future Internet......Page 27
1.3 Internet Congestion Control......Page 29
1.4 Adaptive Congestion Control......Page 33
I: Background on Computer Networks and Congestion Control......Page 38
2.1 Overview......Page 40
2.2.2 Sub-Networks......Page 42
2.2.3 Network Classification......Page 44
2.2.4 LAN Topologies......Page 46
2.3.1 Packet Switching......Page 49
2.3.2 Protocols and Layering......Page 51
2.3.3 Internet Architecture......Page 53
2.3.4 Transfer Control Protocol (TCP)......Page 57
2.3.5 User Datagram Protocol (UDP)......Page 62
2.3.6 Internet Protocol (IP)......Page 63
2.4.1 Queue Size......Page 65
2.4.5 Round Trip Time......Page 66
2.4.6 Fairness......Page 67
2.5 Applications......Page 68
2.5.2 World Wide Web......Page 69
2.5.4 File Transfer......Page 70
2.5.6 Internet Telephony (VOIP)......Page 71
2.6 Concluding Comments......Page 72
3.1 Overview......Page 74
3.2 Core Issues in Congestion Control......Page 75
3.3 TCP: Flow Control and Congestion Control......Page 76
3.3.1 Slow Start......Page 77
3.3.2 Congestion Avoidance......Page 78
3.3.3 Fast Retransmit and Fast Recovery.......Page 80
3.4 TCP Problems......Page 82
3.5.1 TCP Friendliness......Page 84
3.5.2.1 Window-Based vs. Rate-Based......Page 85
3.5.2.2 Unicast vs. Multicast......Page 86
3.5.2.3 End-to-End vs. Router-Based......Page 87
3.6 Concluding Comments......Page 88
4.1 Overview......Page 90
4.2 Drop Tail......Page 91
4.3 Congestion Early Warning......Page 92
4.3.1 Packet Drop Schemes......Page 93
4.3.2 Packet Marking Schemes......Page 97
4.4 Concluding Comments......Page 102
5.1 Overview......Page 104
5.2 Traditional TCP......Page 105
5.3 TCP Modifications for Networks with Large Bandwidth Delay Products......Page 106
5.3.2 HighSpeed TCP (HSTCP)......Page 107
5.3.3 BIC......Page 109
5.3.4 CUBIC......Page 110
5.4 Delay-Based Congestion Control......Page 111
5.4.1 TCP Vegas......Page 112
5.4.2 FAST TCP......Page 113
5.5 Congestion Control for Wireless Networks......Page 114
5.5.1 TCPWestwood......Page 115
5.5.2 TCP Veno......Page 116
5.6.1 Rate Adaptation Protocol (RAP)......Page 117
5.6.2 TFRC......Page 119
5.7 Concluding Comments......Page 120
6.1 Overview......Page 122
6.2 The Fluid Flow Model......Page 123
6.3 Network Representation......Page 124
6.4 Congestion Control as a Resource Allocation Problem......Page 126
6.4.1 Dual Approach......Page 128
6.4.3 Utility Function Selection......Page 129
6.5.1 Stability and Convergence......Page 131
6.5.3 Robustness......Page 132
6.5.4 Fairness......Page 133
6.6 Concluding Comments......Page 134
II: Adaptive Congestion Control Framework......Page 136
7.1 Overview......Page 138
7.2 Packet Switching Network System......Page 139
7.3 Problem Statement......Page 142
7.4 Throughput Improvement......Page 143
7.5 NNRC Framework Description......Page 145
7.5.1 Future Path Congestion Level Estimator......Page 146
7.5.3 Rate Control......Page 147
7.6 Concluding Comments......Page 148
8.2 Feasible Desired Round Trip Time Estimator Design......Page 150
8.2.1 Proof of Lemma 8.1......Page 154
8.2.2 Proof of Lemma 8.2......Page 155
8.3 Rate Control Design......Page 157
8.3.1 Guaranteeing Boundness of Transmission Rate......Page 162
8.3.2 Reducing Rate in Congestion......Page 163
8.4 Illustrative Example......Page 165
8.4.1 Implementation Details......Page 166
8.4.3 Normal Scenario......Page 167
8.4.4 Congestion Avoidance Scenario......Page 168
8.5 Concluding Comments......Page 173
9.2 Necessity for Throughput Control......Page 176
9.3 Problem Definition......Page 178
9.4 Throughput Control Design......Page 179
9.4.1 Guaranteeing Specific Bounds on the Number of Channels......Page 181
9.4.2 Reducing Channels in Congestion......Page 182
9.5 Illustrative Example......Page 183
9.5.1 Implementation Details......Page 184
9.5.2 Normal Scenario......Page 185
9.5.4 Throughput Improvement......Page 188
9.6 Concluding Comments......Page 194
10.1 Overview......Page 196
10.2 Network Topology......Page 197
10.3.1 Effect of Maximum Queue Length......Page 199
10.3.2 Effect of Propagation Delays......Page 201
10.3.3 Effect of Bandwidth......Page 203
10.4 Dynamic Response of NNRC and FAST TCP......Page 205
10.4.1 Bursty Traffic......Page 207
10.4.2 Re-Routing......Page 208
10.4.3 Non-Constant Number of Sources......Page 211
10.5 NNRC and FAST TCP Interfairness......Page 213
10.6 Synopsis of Results......Page 224
10.7 Concluding Comments......Page 225
11.1 Overview......Page 228
11.2.1 QoS Mapping......Page 229
11.2.2 Application QoS Control Design......Page 230
11.3 NNRC Source Enhanced with Application Adaptation......Page 232
11.4 Illustrative Example......Page 233
11.4.1 Application Adaptation Implementation Details......Page 234
11.5 Concluding Comments......Page 235
III: Appendices......Page 240
A.1 Vectors and Matrices......Page 242
A.1.1 Positive Definite Matrices......Page 244
A.2 Signals......Page 246
A.3.1 Continuity......Page 248
A.3.2 Differentiation......Page 249
A.3.3 Convergence......Page 250
A.3.4 Function Properties......Page 251
A.4 Dynamic Systems......Page 252
A.4.1 Stability Definitions......Page 254
A.4.2 Boundedness Definitions......Page 256
A.4.3 Stability Tools......Page 257
B.1 General......Page 272
B.2 Neural Networks Architectures......Page 274
B.2.1 Multilayer Perceptron (MLP)......Page 275
B.2.2 Radial Basis Function Networks (RBF)......Page 277
B.2.3 High-Order Neural Networks (HONN)......Page 278
B.3 Off-Line Training......Page 280
B.3.1.1 Gradient Algorithms......Page 282
B.3.1.2 Least Squares......Page 286
B. 3.1.3 Backpropagation......Page 287
B.4.1 Filtering Schemes......Page 288
B.4.1.1 Filtered Error......Page 289
B.4.1.2 Filtered Regressor......Page 290
B.4.2.2 NLIP Case......Page 291
B.4.3 Steepest Descent Training......Page 292
B.4.4 Recursive Least Squares Training......Page 293
B.4.5 Robust On-Line Training......Page 294
Bibliography......Page 298
Back Cover......Page 326
